Re: [PATCH v3 1/2] dt-bindings: hwmon: Add MAX6639

[Rob Herring]

On Fri, Aug 04, 2023 at 09:10:37AM -0700, Guenter Roeck wrote:
> On 8/4/23 08:48, Conor Dooley wrote:
> > On Thu, Aug 03, 2023 at 04:43:59PM +0200, Naresh Solanki wrote:
> > > From: Marcello Sylvester Bauer <sylv@xxxxxxx>
> > > 
> > > Add binding documentation for Maxim MAX6639 fan-speed controller.
> > > 
> > > Signed-off-by: Marcello Sylvester Bauer <sylv@xxxxxxx>
> > > Signed-off-by: Naresh Solanki <Naresh.Solanki@xxxxxxxxxxxxx>
> > > ---
> > > Changes in V3:
> > > - Update title
> > > - Add pulses-per-revolution, supplies & interrupts
> > > Changes in V2:
> > > - Update subject
> > > - Drop blank lines
> > > ---
> > >   .../bindings/hwmon/maxim,max6639.yaml         | 60 +++++++++++++++++++
> > >   1 file changed, 60 insertions(+)
> > >   create mode 100644 Documentation/devicetree/bindings/hwmon/maxim,max6639.yaml
> > > 
> > > diff --git a/Documentation/devicetree/bindings/hwmon/maxim,max6639.yaml b/Documentation/devicetree/bindings/hwmon/maxim,max6639.yaml
> > > new file mode 100644
> > > index 000000000000..b3292061ca58
> > > --- /dev/null
> > > +++ b/Documentation/devicetree/bindings/hwmon/maxim,max6639.yaml
> > > @@ -0,0 +1,60 @@
> > > +# SPDX-License-Identifier: (GPL-2.0 OR BSD-2-Clause)
> > > +%YAML 1.2
> > > +---
> > > +$id: http://devicetree.org/schemas/hwmon/maxim,max6639.yaml#
> > > +$schema: http://devicetree.org/meta-schemas/core.yaml#
> > > +
> > > +title: Maxim MAX6639 Fan Controller
> > > +
> > > +maintainers:
> > > +  - Naresh Solanki <Naresh.Solanki@xxxxxxxxxxxxx>
> > > +
> > > +description: |
> > > +  The MAX6639 is a 2-channel temperature monitor with dual, automatic, PWM
> > > +  fan-speed controller.  It monitors its own temperature and one external
> > > +  diode-connected transistor or the temperatures of two external diode-connected
> > > +  transistors, typically available in CPUs, FPGAs, or GPUs.
> > 
> > > +  fan-supply:
> > > +    description: Phandle to the regulator that provides power to the fan.
> > 
> > > +  pulses-per-revolution:
> > > +    description:
> > > +      Define the number of pulses per fan revolution for each tachometer
> > > +      input as an integer.
> > > +    $ref: /schemas/types.yaml#/definitions/uint32
> > > +    enum: [1, 2, 3, 4]
> > > +    default: 2
> > 
> > Apologies if I am digging up old wounds here, since there was quite a
> > bit of back and forth on the last version, but these two newly added
> > properties look to be common with the "pwm-fan" and with
> > "adi,axi-fan-control". At what point should these live in a common
> > schema instead?
> > 
> > Otherwise, this looks okay to me, although I'll leave things to
> > Krzysztof since he had a lot to say about the previous version.
> > 
> 
> Rob has said that he won't accept any fan controller bindings without a generic
> schema. At the same time he has said that he expects properties such as the
> number of pulses per revolution to be attached to a 'fan' description, and he
> wants pwm related properties of fan controllers to be modeled as pwm controllers.
> And now we have a notion of a regulator providing power to the fan (which again
> would be the fan controller, at least in cases where the fan controller
> provides direct voltage to the fan). On top of that, this fan-supply property
> should presumably, again, be part of a fan description and not be part of the
> controller description. I don't think anyone knows how to make this all work
> (I for sure don't), so it is very unlikely we'll see a generic fan controller
> schema anytime soon.

I thought what was done earlier in this series was somewhat close. And 
there are some bindings that already look pretty close to what a common 
binding should. But it seems no one wants to worry about more than their 
1 device.

In case it's not clear, as-is, this binding is a NAK for me.

> Given that neither fan-supply nor pulses-per-revolution is implemented in the
> driver, and given that I am not aware of any fans which would have a value for
> pulses-per-revolution other than 2, my personal suggestion would be to add the
> chip to trivial devices and be done with it for the time being.

I'm fine with that too. Just keep kicking that can...

Rob